New Frontiers in Practical Risk Management

English edition Issue n. 8 - Fall 2015

Iason ltd. and Energisk.org are the editors of Argo newsletter. Iason is the publisher. No one is al-lowed to reproduce or transmit any part of this document in any form or by any means, electronicor mechanical, including photocopying and recording, for any purpose without the express writtenpermission of Iason ltd. Neither editor is responsible for any consequence directly or indirectly stem-ming from the use of any kind of adoption of the methods, models, and ideas appearing in the con-tributions contained in Argo newsletter, nor they assume any responsibility related to the appropri-ateness and/or truth of numbers, figures, and statements expressed by authors of those contributions.

New Frontiers in Practical Risk ManagementYear 2 - Issue Number 8 - Fall 2015

Published in January 2016First published in October 2013

Last published issues are available online:www.iasonltd.comwww.energisk.org

Fall 2015

NEW FRONTIERS IN PRACTICAL RISK MANAGEMENT

Editors:Antonio CASTAGNA (Co-founder of Iason ltd and CEO of Iason Italia srl)Andrea RONCORONI (ESSEC Business School, Paris)

Executive Editor:Luca OLIVO (Iason ltd)

Scientific Editorial Board:Fred Espen BENTH (University of Oslo)Alvaro CARTEA (University College London)Antonio CASTAGNA (Co-founder of Iason ltd and CEO of Iason Italia srl)Mark CUMMINS (Dublin City University Business School)Gianluca FUSAI (Cass Business School, London)Sebastian JAIMUNGAL (University of Toronto)Fabio MERCURIO (Bloomberg LP)Andrea RONCORONI (ESSEC Business School, Paris)Rafal WERON (Wroclaw University of Technology)

Iason ltdRegistered Address:6 O’Curry StreetLimerick 4Ireland

Italian Address:Piazza 4 Novembre, 620124 MilanoItaly

Contact Information:info@iasonltd.comwww.iasonltd.com

Energisk.orgContact Information:contact@energisk.orgwww.energisk.org

Iason ltd and Energisk.org are registered trademark.

Articles submission guidelinesArgo welcomes the submission of articles on topical subjects related to the risk management. Thetwo core sections are Banking and Finance and Energy and Commodity Finance. Within these twomacro areas, articles can be indicatively, but not exhaustively, related to models and methodologiesfor market, credit, liquidity risk management, valuation of derivatives, asset management, tradingstrategies, statistical analysis of market data and technology in the financial industry. All articlesshould contain references to previous literature. The primary criteria for publishing a paper are itsquality and importance to the field of finance, without undue regard to its technical difficulty. Argois a single blind refereed magazine: articles are sent with author details to the Scientific Committeefor peer review. The first editorial decision is rendered at the latest within 60 days after receipt of thesubmission. The author(s) may be requested to revise the article. The editors decide to reject or acceptthe submitted article. Submissions should be sent to the technical team (argo@iasonltd.com). LaTex orWord are the preferred format, but PDFs are accepted if submitted with LaTeX code or a Word file ofthe text. There is no maximum limit, but recommended length is about 4,000 words. If needed, forediting considerations, the technical team may ask the author(s) to cut the article.

ADVERTISING FEATURE

chasing the future in Finance

Iason ltd provides solutions that enable managers, risk-control officers and front office peopleto understand and value their portfolios. All solutions are built on the principles of soundness,accuracy and clearness.

PRICING

Modern Derivatives PricingSound models to value derivative instrumentsaccording with the best and newest marketpractice (CVA, DVA and liquidity).

Complex ProductsSophisticated analytical models, supported byresearch, in order to effectively and efficientlyevaluate all kind of exotic structures,accounting for all the relevant risk factors.

Revaluation of BooksThe service is for those institutions notinterested in building a trading and riskmanagement system, but simply needing a fairrevaluation of their (possibly very complex)derivatives positions.

RISK MEASUREMENT

Credit RiskInnovative approach to Credit VaR (CreVar)modeling and computation.

Market RiskInnovative and sophisticated solutions formarket risk measure and management.

Counterparty RiskPioneers in the development of CCR and CVAmeasurement models and applications withdevoted internal teams in many major banks.

ALM and Liquidity RiskConsulting services to support ALM andtreasury department of financial institutions.Innovative models taking into account theinterrelation between funding, liquidity andCCR.

To learn more: www.iasonltd.com.To contact us: info@iasonltd.com. Follow us on:

@Iason_ltd company web page iason iason.network

3

NEW FRONTIERS IN PRACTICAL RISK MANAGEMENT

Table of Contents

Editorial pag. 05

Antonio Castagna, Andrea Roncoroni and Luca Olivo introduce the new topics of this n. 8 Argo edition.

banking & finance

The Revision of the CVACapital Charge by the Basel Commitee pag. 07

Antonio Castagna and Michele Bonollo

energy & commodity finance

When has OPECSpare Capacity Mattered for Oil Prices? pag. 15

Hilary Till

Front Cover: Luigi Russolo Dinamismo di un’automobile, 1912-13.

EDITORIAL

Dear Readers,

The Fall 2015 issue of Argo magazine represents an importantstep in the life-cycle of the newsletter. It marks two years of publica-tions indeed, that have touched several topics in the financial fields,from the collateral management to the most advanced models inderivative pricing, passing through the ultimate news in regulation.We really hope to keep going on in this direction in the next years.

The magazine starts with an interesting review in theBanking & Finance section. Antonio Castagna and MicheleBonollo introduce the recent BCBS paper containing a deepreview of the CVA capital charge. In the article, theyhighlight the main features of this new framework to-gether with the identification of some potential drawbacks.

In the Energy and Commodity Finance part, another inter-esting contribution by Hilary Till is presented. This timethe focus is on oil market: the author argues that, dur-ing some particular periods of time, the OPEC spare ca-pacity is the most important factor for driving oil prices.

We conclude as usual by encouraging the submission ofcontributions for the next issues of Argo in order to im-prove each time this newsletter. Detailed information aboutthe process is indicated at the beginning. New and challeng-ing articles are upcoming with the next releases indeed.

Enjoy your reading!

Antonio CastagnaAndrea Roncoroni

Luca Olivo

Fall 20155

NEW FRONTIERS IN PRACTICAL RISK MANAGEMENT

AAAA

Banking & Finance

CVA Regulation

6

The Revision to theCVA Capital Charge bythe Basel CommitteeShort Review and some critical Issues

With the recent BCBS paper no.325 a newconsultative session started, concerninga deep review to the CVA capital charge,that was introduced with the CRR (Basel3) regulation. This revision aims to im-prove the CVA calculation with respect tothe forthcoming fundamental review oftrading book (FRTB), and to have a morerisk sensitive approach for the standardapproach. Nevertheless, in this first for-mulation some points are not well clar-ified or solved. In this article a shortreview is presented, focusing on somehighlights about the new framework to-gether with the identification of somedrawbacks of the suggested new set up.

Antonio CASTAGNAMichele BONOLLO

In the Basel II regulation most of the effort wasdevoted to set the internal model for the creditrisk and (IRB) and to define the new capital

charge for the Operational Risk.Despite the relevant improvement to the old

Basel I regulation , the regulation was still weak fortwo main reasons

• Some building blocks, e.g. risk sources to befaced by the banks with their own capitals,were still forgotten.

• Some discrepancies (or arbitrage tricks) werenot removed. More explicitly, the inclusion ofa position in the banking book vs. the tradingbook or its asset type (e.g. bond vs derivative) al-lowed some “unfair” differences in the capitalrequirement calculation.

For this reason the Basel 2.5 and Basel 3 reg-ulations attempted to solve the above issues byintroducing the new IRC (Incremental Risk Charge)and CVA (Credit Value Adjustment) building blocks.See [2], [3], [4] and [5].

The IRC, mainly for the bond positions, impliesa capital charge for the downgrade and defaultrisk. The CVA requires a capital requirement forthe positions in OTC derivatives due to counterpartydowngrade. In this way the huge unrealized (andunexpected losses) of the derivatives positions mustbe covered by the capital. After the 2008 crisis manybanks were closed to default because the own fundswere not enough to face this kind of losses. Let usrecall that the fair value principle of the accountingstandards obliges to include in the book evaluationalso the counterparty credit quality.

Finally, by these innovations, bonds and OTCderivatives are subject to the same building blocks:

• Default risk, respectively IRC and CCR

• Downgrade risk, IRC and CVA

Fall 20157

CVA REGULATION

• Market risk.

The CVA current review. Purposes andContents

In July 2015 the Basel Committee issued the consul-tative paper 325, see [1], concerning the revision ofthe credit value adjustment.

The main purposes can be summarized as fol-lows:

1. To capture all the risk factors in the CVA cap-ital charge. The most relevant point here is totake in to account also the market risk factorsthat can affect the future value of the expo-sure. This is also due to the practices in themarket, where also the CVA hedging dealsare sensitive to these market factors

2. To be compliant with the accounting princi-ples. More specifically, the future exposureshould be calculated with a market implied cal-ibration of the parameters (i.e. underlyingvolatility) with a risk neutral approach. Thisis a strong new perspective, as currently inthe CCR framework for the EAD estimationin the internal models the market calibrationis not required.

3. Alignment of the CVA to the new marketrisk framework. We refer mainly to the use ofsensitivities, risk factors taxonomy and so on.

4. A more risk sensitive approach for the noninternal calculations.

A 3-levels hierarchy of possible approaches wasstated in this draft version of the paper, namely:

• FRTB-CVA framework, that splits in:

A IMA-CVA = Internal model approach

B SA-CVA = Standardized approach

• Basic CVA framework

The last approach is eligible for the banks thatare not able to match the FRTB requirements orhave not enough resources to implement such aproject.

The new framework critical points

We contributed to the first consultative session, see[6]. In this section we highlight a set of issues thatin our opinion are not solved in a satisfactory way.

The CVA Hierarchy

We appreciate the effort of the Committee to dif-ferentiate the approach for the CVA capital charge,in order to take in to account a proportionality ap-proach, section 2 of the Consultative paper. Here forproportionality we mean both the complexity/sizeof banks OTC book and their internal capabilities.

Nevertheless, we do not understand completelythe rationales for the suggested hierarchy, mainlyreferred to the Basic Approach. In fact:

• To compute a rough sensitivity with respectto the spread of the counterparty is not sodifficult. Many banks adapt in some way forsuch a task their pricing libraries.

• Moreover, small and medium banks very of-ten have plain interest derivatives with theircorporate customers. Hence, the future dy-namics of the exposure and the interactionbetween credit risk factors and market factorsmay be faced by usual mathematical tools

• On the other hand, the very hard problemis not the sensitivity, but the sensitivity withrespect to what. In other words, the counterpar-ties may be illiquid, and the proxy of it to anyindex is still an open issue. The section C ofthe Annex, point 103, addresses only partiallythis issue, since it is just a judgmental calibra-tion of the involved market parameters. An“implied” market calibration is very difficultto perform.

• Finally, the basic approach seems not to becoherent with the recent strategy for the othercapital charge building blocks. We recall thatin the new standardised approaches for theCCR (paper 279) and for the market risk FRTB(paper 305), the goal is to have more effectivemodels, i.e. to make them more risk sensitiveand more “granular” in the risk factors treat-ment. Also for the Operational Risk (paper291) the “basic” approach, i.e. the BIA basedon just the gross income risk driver, will bereplaced by a new more sophisticated set ofexposure indicators.

To summarize, for the above specific reasonsand for a coherency principle, we find that theCommittee should wonder about the taxonomy, inparticular on the effective needing of a basic ap-proach.

8iasonltd.com

Regulatory vs. Accounting standard

The consultative paper definition of the CVA triesto fill the existing gap between the regulatory andaccounting CVA. This will lead to a greater con-sistency of the risk assessment determined by theBasel rules and the accounting measuring of theCVA metric.

One of the differences that will remain, evenwith the new framework proposed in the paper, isthe treatment of the DVA (see par. 1 and 13), whichis excluded in agreement with all the most recentBasel regulation, and which is on the contrary recog-nised by the new accounting principles, namelyIFRS 13. We recall that the DVA, debt value adjust-ment, allows to the banks to apply an adjustment(to decrease the fair value) to the debt positions (i.e.OTC derivatives in sell side of the portfolio) due tothe credit quality their own credit quality.

We believe that the current regulation is follow-ing the right route in excluding from the CVA theDVA, i.e. in considering only the unilateral CVA:we think this is right not only under a prudentialpoint of view, but also under a sound financialperspective. Actually, it has been proved in someresearch papers (see for example Castagna [9], [10])that the DVA can be replicated under very strictconditions hard (although not impossible in princi-ple) to implement in practice. But even concedingthe actual possibility to replicate the DVA, it can beshown that its accounting at a counterparty netting-set level largely over-estimate the limited liabilityprotection that shareholders have.

Hence, the choice to exclude the DVA is reason-able but it is not just prudent, contrasting the ac-counting principles laid down in the IFRS 9. Theseprinciples rely on the false premise that a derivativecontract has an “objective” value that is indepen-dent from which of the two counterparties it isevaluated, even if this objective value contains ele-ments that belong to both counterparties, namelythe adjustments for the credit risks of both of them.This premise paves the way to an exact symmetricalevaluation of the two parties that is quite nice un-der an accounting point of view, since it makes theevaluation principles in theory perfectly complyingwith general accounting principle of the “fair andtrue view”, but it can be hardly justified as far asthe other general principle of “prudence”.

We think that both the regulatory and account-ing prescriptions can be fulfilled if banks allocatea provision equal to the DVA any time a variationof the DVA is added to valuation of the derivativeportfolio. In this case, all the gains coming fromthe DVA’s increase would be neutralized, since theywould be offset by a parallel increase of the provi-

sion, and the other hand when the DVA collapsesto zero as the trade expires, the loss would be com-pensated by a reduction of the reserve. In the end,the DVA would be fairly represented in the valu-ation of the derivative portfolio, yet it would notadd to the bank P&L and it would be recognizedsimply as a cost at inception and gradually split onthe years of the portfolio’s duration. The allocationof the provision does not require any change inthe accounting principles, because it adheres in anycase to the overarching “prudence” principle.

Banks would be incentivized not to start anyhedging activity of the DVA, since the allocationof the provision would create P&L volatility if thebank tried to start a hedge. The possible complaintsthat the regulation does not match the accountingprinciples would thus be addressed.

Specific calculation methodology

The paper seems to propose specific calculationmethodologies for sensitivities (see par. 41 and 42in [1]). Actually, more sophisticated methods tocompute sensitivities, other than the brute force ap-proach proposed, are available in theory and theyhave also been implemented in practice by somebanks (e.g.: adjoints and automatic differentiation).Simple bumping the risk factors may not be themost effective way to compute sensitivities. See [8].

We suggest modifying the CVA framework toallow banks to choose the preferred numerical cal-culation method. In other words, we believe thatthe Committee should strictly prescribe only thefunctional mathematical definition of the indicator,allowing to the bank to select the optimal strat-egy to calculate it (numerical, simulation, etc.). SeeBonollo et al. [7].

A similar issue arises in the EEt calculation forthe EPE in CCR, where the theoretical definition(expected value in the future) is combined with thealgorithm (a Montecarlo approach). We fear that thisway to state the regulation could be misunderstood,since in practice the banks choose their own algo-rithmic strategy, combining ICT devices (GPU, grid,etc.) with mathematical tools (quasi Montecarlo,approximations, etc.).

Non-Captured Risks

The CD suggest a multiplier mCVA for the wrong wayrisk, if it is not properly accounted for in the bank’smethodology (see par. 32 and 33). Now, while weagree with a greater prudence in the assessment ofthe CVA for the un-accounted risks, we think thatthe CD focuses only on one of them, namely: thewrong way risk, without considering more relevant

Fall 20159

CVA REGULATION

risks likely affecting the measurement in a morematerial way.

More explicitly the CD does not mention the er-rors due to the correlation matrix employed in thecalculations, when it does not reflect the actual fu-ture matrix. Correlations are rarely found quoted inthe market and only a few contracts can be tradedto hedge the correlation risk. Since the CVA, asfar as its hedging is considered, can be seen as avery complex hybrid derivative contract (especiallyif netting sets are cross-asset), a wrong correlationmatrix implies that the second order sensitivities(Cross-Gamma and Cross-Vega) cannot be soundlyhedged, and this would entail a mis-hedge also forthe linear Greeks (i.e.: Delta and Vega) which mayeventually result in a global increase of the P&Lvolatility, contrarily to the supposed minimizationdue to the hedging.

We suggest including in the CVA frameworkalso an assessment of the risks related to the cor-relation matrix, when correlations cannot be easilytraded in the market, or they cannot be traded atall. This is very likely the most common situationin the current markets.

As a general consideration, we doubt aboutthe effectiveness of the CVA hedging (i.e.: repli-cation) in practice. We would prefer to treat theCVA earned on the deals closed by the bank as anactuarial premium, rather than a derivative expo-sure to be synthetically replicated. The volatility ofthe CVA can surely absorb capital but the effective-ness of the hedging, set up to reduce it, should becarefully evaluated and put under stress.

FRTB CVA and Double Counting

The CVA framework relies on the changes of theFundamental Review of the Trading Book, yet itis quite independent from the calculation of theregulatory capital for the market risk. We are awarethat the Expected Shortfalls (ES) for the market riskand for the CVA volatility are computed out of twoquite different approaches, and that the latter needsa simulation up to the expiry of the longest contractwhich is not strictly needed to determine the mar-ket ES. Nonetheless, it is obvious that the positiveexposures increasing the counterparty risk could bealso compensating a decreasing risk on the marketrisk side, due to the positive impact on the NPVs.

The CVA framework should allow banks withsophisticated skills and strong IT computationalcapabilities to measure jointly the ES on the marketand counterparty risks, so that possible compensa-tion of risks are properly identified and measured.

The Risk Measure and Aggregation in the FRTBSA-CVA

The computational workflow.The SA approach follows the general strategy ofmaking the standardized model more risk sensitive,by dealing in a rigorous way the key concepts suchas risk factors, sensitivities, dependency/correlationstructure. Generally the parameters implied by thisset-up are assigned by the Committee.

All the non-internal-models banks will be obligedto work intensively to switch from very simplestandard models that do not require sophisticatedmapping and calculation procedures to the new SA.We refer mainly to Market, CCR and CVA capitalcharge.

In this framework, we think that the SA modelsshould be more homogenous in their “architecture”,to avoid that small-medium banks make some con-fusion in managing and calculating these new mea-sures. Otherwise they are obliged to maintain atthe same time two or more systems for mappingand categorizing their risk factors.

As a simple example, let us compare the newSA-EAD for the CCR (paper 279) with the currentSA-CVA. For the sake of simplicity, we refer brieflyto the “Equity” asset class:

• For CCR (EAD) purposes, a single risk factormodel is prescribed with just one hedging set.Hence the full offset is allowed within thesame reference entity, while a correlation fac-tor is assigned with respect to the systematicfactor, 50% for single names and 80% for theindices

• For the CVA, i.e.: for capturing the (equity) ex-posure volatility, the workflow consists of 10buckets given by a sector/geography taxonomy.The Delta and Vega exposure CVA sensitivi-ties are then calculated with a cross correla-tion ρij of 15% between all couples of buckets.

Then we could easily build a counterexample,e.g.: a portfolio of 2 equity derivatives belonging todifferent buckets, where the price joint movementsof the 2 entities are taken in to account differentlyfor the PFE and for the CVA-Exposure effect respec-tively.

Correlation coefficients and Risk Weights.In the spirit of the SA calculation, we generallyagree with the general workflow. On the otherhand, we suggest improving some of the currentparameters value. There are several examples, butfor brevity we point out just some simple cases:

10iasonltd.com

• The risk weight RW for Delta risk for FX is15%, while the lowest RW for the Equity riskclass is 30%. As well known, the RW roleis to move from a what-if measure (the sen-sitivity of the instrument to the risk factor)to the instrument volatility. From this set-upone could argue that the highest volatility FXrate is 2 times lower than the lowest equityvolatility. We claim that this is not a realisticpicture of the market price volatilities

• The correlation cross buckets for the FX is 0.6,for the Equity is 0.15. Again, we find it notvery accurate. It is often observed in the finan-cial markets that the sectors move together in-side a macro area, irrespectively of the size ofthe firms. In some cases, such as the buckets(1,2,3,4) vs bucket (9), i.e.: large cap vs. smallcap in emerging markets (see Annex 1.B.2), a0.15 coefficient is too low and not conserva-tive. On the other hand, a 0.6 “flat” betweenthe currencies in some cases is too high, alsoin a conservative perspective.

Computational Burden for the FRTB IMA-CVA

The CD proposes to compute the CVA capitalcharge daily, under different assumptions on theMPOR (par. 13), liquidity horizons (par. 85) and,above all, separately and jointly for all the relevantrisk factors (par. 86, 87, 88, 89). On the one hand,this is a huge increase of the computational bur-den with respect to the current requirements to

calculate the IRC; on the other hand, even moresophisticated banks calculating the CVA for thetrading desks managing it, likely do not operatedaily so many computations as those implied in theproposal.

IT technology is certainly available to performthe required computations on a daily frequency, butwe suspect that the investments needed to upgradeexisting systems would be massive even for moreadvanced institutions.

We are not trying to minimize the complexityand the subtleties of the risks involved with theCVA (as the point above on the correlation shows).We want simply to point out that a daily calcula-tion represents a too high a frequency for most ofpractical purposes. In our view, it would be betterto relax the frequency in favour of a deeper analy-sis of the model risks even beyond those explicitlyconsidered in the CD.

ABOUT THE AUTHORS

Antonio Castagna is Senior Consultant, co-founder ofIason ltd and CEO at Iason Italia srl.Email address: antonio.castagna@iasonltd.com

Michele Bonollo is Senior Consultant at Iason Ltd andIMT Lucca.Email address: michele.bonollo@iasonltd.com

ABOUT THE ARTICLE

Submitted: November 2015.

Accepted: December 2015.

References

[1] Basel Committee on Banking Supervision. Review ofthe Credit Valuation Adjustment Risk Framework. 2015.Available at http://www.bis.org/bcbs/publ/d325.htm

[2] Basel Committee on Banking Supervision.Basel III Summary Table. 2014. Available athttp://www.bis.org/bcbs/basel3/b3summarytable.pdf

[3] Basel Committee on Banking Supervision. BaselIII Phase-In Arrangements. 2014. Available athttp://www.bis.org/bcbs/basel3/basel3phaseinarrangements.pd f

[4] Basel Committee on Banking Supervision. Guide-lines for computing capital for incrementalrisk in the trading book. 2009. Available athttp://www.bis.org/publ/bcbs159.htm

[5] EU. Regulation (EU) No 575/2013 on prudential re-quirements for credit institutions and investmentfirms. 2013. Available at http://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:32013R0575

[6] Bonollo M. and A. Castagna. Comments to theConsultative paper Review of the Credit Valua-tion Adjustment Framework. 2015. Available athttp://www.bis.org/bcbs/publ/comments/d325/iason.pdf

[7] Bonollo M., L. Di Persio, I. Oliva and A. Semmoloni. AQuantization Approach to the Counterparty Credit Expo-sure Estimation. 2015. Available at www.ssrn.com

[8] Capriotti L. Fast Greeks by Algorithmic Differentiation.Journal of Computational Finance, 14 (3), pp. 3-35. 2011.

[9] Castagna, A. On the dynamic replication ofthe DVA: Do banks hedge their debit value ad-justment or their destroying value adjustment?.2012. Available at http://www.iasonltd.com/wp-content/uploads/2013/02/1d.pdf

[10] Castagna, A. On the impossibility to replicate the DVA.Risk. November, 2012.

Fall 201511

NEW FRONTIERS IN PRACTICAL RISK MANAGEMENT

AAAA

Energy & CommodityFinance

Crude Oil Market

12

ADVERTISING FEATURE

http://energy-commodity-finance.essec.edu

http://energy-commodity-finance.essec.edu/events/monthly-executive-seminar

Fall 201513

ADVERTISING FEATURE

J.P. Morgan Center for Commodities and University of Colorado Denver Commodities

Research Fellowship Awards

The J.P. Morgan Center for Commodities (JPMCC) at the University of Colorado Denver is pleased to announce 3 (three) Research Fellowship Awards to support

applied commodities research projects. Each Fellowship Award is for $8,000, plus travel costs to Denver (for the principal author). The Awards will be disbursed on a

staggered basis, with the expectation that a completed and presentable working paper be ready by January 2017. The selected Fellows will be required to present their findings at a research forum hosted by JPMCC. The working papers will be posted on the JPMCC's web site. The Fellowships have been funded by the CME

Group Foundation. Ph.D. students and early career professionals are stronglyencouraged to apply.

PAPER SUBMISSION PROCEDUREProposals should be sent electronically as one packet, with the subject lineCommodities Research Fellowship Award application, tocommodities.center@ucdenver.edu, and contain all three of the following materials: acover page indicating the significance of the research, a description of the proposal (atmost ten double spaced pages including graphs, tables and references), and theapplicant's current curriculum vitae.Deadline for receipt of applications is February 28, 2016.

J.P. MORGAN CENTER FOR COMMODITIESThe JPMCC conducts research and publishes results to enhance the understanding ofcommodities in academia, commodities industries, policy, media and the generalpublic. The JPMCC's recently formed Research Council, to help shape the researchagenda of the JPMCC, consists of eminent researchers and successful professionals ofpractice, interested in a shared platform for rigorous and relevant commoditiesresearch and its dissemination to the wider commodities community. For moreinformation on the JPMCC's activities, please visit its website at: http://goo.gl/F5T27Y

SELECTION CRITERIAThe JPMCC Research Council's Fellowships Grant Committee will select proposals thatbest align with the curriculum vitae will be the two main sources of information usedto evaluate the significance of the research. The proposed research will be judged bythe quality and significance, the proposed method and analysis, its degree of originalityand rigor, and the likelihood that the proposals will have practical application in thefield of commodities and result in peer reviewed publications. Eligibility is not confinedto University faculty; both academic researchers and industry professionals areencouraged to apply. Completed projects are not eligible.

Awards will be announced before summer 2016. For further information please address any question to commodities.center@ucdenver.edu

14

When has OPEC SpareCapacity Mattered for OilPrices?

Oil prices usually “feed off multiple in-fluences,” as noted in Büyüksahin (2011).The various influences on oil prices are il-lustrated in Figure 1. But are there timeswhen OPEC spare capacity is the mostimportant factor for driving oil prices?This article will argue the answer isyes, and will discuss the circumstanceswhen this has been the case in the past.

Hilary TILL1

The current definition of spare capacity isas follows. The U.S. Energy InformationAdministration (EIA) has defined “sparecapacity as the volume of production that can

be brought on within 30 days and sustained for at least90 days. [. . . ] OPEC spare capacity has provided anindicator of the world oil market’s ability to respond topotential crises that reduce oil supplies,” according toEIA (2014).

OPEC Spare Capacity Mattered in 2008

As discussed in Till (2015), to motivate why thespare capacity situation might be quite importantto the behavior of crude oil prices, one can reviewthe circumstances of 2008. We found out from theevents of that year what can happen if the oil excess-capacity cushion becomes quite small. In July 2008,

the role of the spot price of oil was arguably to finda level that would bring about sufficient demanddestruction, after which the spot price of oil spec-tacularly dropped. This explanation is drawn fromresearchers from both the Federal Reserve Bank ofDallas and the U.S. Commodity Futures TradingCommission.

Figure 2 excerpts from a Federal Reserve Bankof Dallas paper. The red line shows WTI priceswhile the blue line is OPEC excess capacity. WhenOPEC excess capacity levels reached pinch-pointlevels, the price of crude oil responded by explod-ing.

Figure 3 provides another way of illustratingwhat happened to the price of crude oil as OPECspare capacity collapsed in mid-2008. It shows WTIoil prices on the y-axis and OPEC spare capacity onthe x-axis. The dark blue dots are data-points fromJanuary 1995 to February 2004, while the pink dotsare from March 2004 to August 2008, as OPEC sparecapacity became ever lower. This graph is analo-gous to the typical economics-of-storage graph, asconceptually illustrated in Figure 4, where the priceof a commodity can become exponentially highwhen there are low enough inventories. In the caseof crude oil, though, the relevant variable on thex-axis had been spare capacity over the timeframerepresented by Figure 3.

1The work leading to this article was jointly developed with Joseph Eagleeye of Premia Research LLC. Research assistance fromKatherine Farren, CAIA, of Premia Risk Consultancy, Inc. is gratefully acknowledged.

Fall 201515

CRUDE OIL MARKET

FIGURE 1: Diagram based on Büyüksahin (2011).

Structural Break after 2008

In Büyüksahin (2011), the energy researcher showsthat the relationship illustrated in Figure 3 struc-turally changed. This point is illustrated in Figure5 with the addition of data from September 2008through September 2015; these data-points are inlight blue. Using data through September 2015, itis not clear what the relationship between WTI oilprices and OPEC spare capacity is, if any.

More recently, Kibsgaard (2015) also pointed outthat the established relationship between oil pricesand OPEC spare capacity had broken down, as il-lustrated in Figure 6. In this particular graph, Brentprices are used instead of WTI prices and are repre-sented by the blue line while OPEC spare capacityis presented in terms of percentage-of-global-oil-demand and is represented by the green line.

When Has OPEC Spare Capacity Mat-tered?

We can conclude from the previous section thatit may only be in a certain state-of-the-world thatOPEC spare capacity matters. But what preciselydescribes that particular state-of-the-world?

Ori (2015) essentially provides the answer.OPEC spare capacity should only matter if one is ina state of low inventories. Figure 7 shows how lowlevels of current and expected OPEC spare capacityare mirrored by increases in current and expectedglobal crude oil inventories.

We can now re-examine Figure 5 based on Ori(2015)’s insight. Let us examine the relationshipbetween WTI oil prices and OPEC spare capacityfrom January 1995 through September 2015, butonly when crude oil inventories are low. We will checkif there might be a clear relationship using U.S. oilinventories. This particular conditional examina-tion is illustrated in Figure 8. At least over theperiod, January 1995 through September 2015, it isapparent that tight levels of OPEC spare capacityhad only mattered when U.S. oil inventories werelow. Here, we define low levels of inventories as be-ing under 22.4 days-of-forward-supply-of-crude-oilin the U.S.

Economics of Price Volatility for CrudeOil

Harrington (2005) would not be surprised by Figure8. This author noted that the true buffer againstcrude-oil price shocks should be represented as notjust above-ground stocks, but also spare-productioncapacity. In the absence of being able to draw oninventories or exploit surplus capacity, price is theonly lever that can balance supply-and-demand insuch a scenario.

We can now note the conditions under whichthe generalized economics-of-price-volatility dia-gram shown in Figure 4 may apply to crude oil:when inventories are sufficiently low, decreasingOPEC spare capacity has produced the same pat-tern as in this conceptual diagram.

16energisk.org

FIGURE 2: Graph based on Plante and Yücel (2011), Chart 2. The red line is WTI prices while the blue line is OPEC excess capacity.Oil prices are monthly averages. Sources of Data: U.S. Energy Information Administration (EIA) and the Wall Street Journal.

FIGURE 3: Graph updated from Till (2014), Slide 19. Sources of Data: the WTI Spot Price is the "Bloomberg West Texas Interme-diate Cushing Crude Oil Spot Price". The following Bloomberg formula was used to create a monthly data set from daily prices:bdh("USCRWTIC Index","px last","1/1/1995","8/31/2008","per=cm","quote=g"). The OPEC Spare Capacity data is from the U.S. En-ergy Information Administration’s website, which was accessed on 8/30/14 (for the 1995 data) and on 10/24/15 (for the 1996 throughSeptember 2015 data.). Presenting data in this fashion is based on Büyüksahin et al. (2008), Figure 10, which has a similar, but notidentical, graph. Their graph, instead, shows “Non-Saudi crude oil spare production capacity” on the x-axis.

Fall 201517

CRUDE OIL MARKET

FIGURE 4: Diagram based on Wright (2011), Slide 39.

FIGURE 5: Sources of Data: The WTI Spot Price is the "Bloomberg West Texas Intermediate Cushing Crude Oil Spot Price".The following Bloomberg formula was used to create a monthly data set from daily prices: bdh("USCRWTIC Index","pxlast","1/1/1995","9/30/2015","per=cm","quote=g"). The OPEC Spare Capacity data is from the U.S. Energy Information Admin-istration’s website, which was accessed on 8/30/14 (for the 1995 data) and on 10/24/15 (for the 1996 through September 2015 data).Presenting data in this fashion is based on Büyüksahin (2011), Slide 49, which has a similar, but not identical, graph. His graph,instead, shows “Non-Saudi crude oil spare production capacity” on the x-axis and is updated through August 2010.

18energisk.org

FIGURE 6: Graph based on Kibsgaard (2015), Slide 5. Sources of Data: International Energy Agency and Schlumberger Analytics.

FIGURE 7: Charts based on Ori (2015). Source of Data: EIA.

Fall 201519

CRUDE OIL MARKET

FIGURE 8: Sources of Data: The WTI Spot Price is the "Bloomberg West Texas Intermediate Cushing Crude Oil SpotPrice". The following Bloomberg formula was used to create a monthly data set from daily prices: bdh("USCRWTIC In-dex","px last","1/1/1995","9/30/2015","per=cm","quote=g"). The OPEC Spare Capacity data is from the U.S. Energy Informa-tion Administration’s website, which was accessed on 8/30/14 (for the 1995 data) and on 10/24/15 (for the 1996 throughSeptember 2015 data). “Days Forward Supply” refers to the U.S. Department of Energy’s U.S. Days-of-Supply-for-Crude-Oil.The following Bloomberg formula was used to create a monthly data set from weekly data: bdh("DSUPCRUD Index","pxlast","1/1/1995","9/30/2015","per=cm","quote=g"). Presenting data in this fashion is based on Büyüksahin et al. (2008) andBüyüksahin (2011).

Caveats

Now, a careful reader may note a particular empha-sis on OPEC spare capacity, ignoring non-OPECproducers. According to IMF (2005), “non-OPECproducers do not have the incentive to maintainspare capacity as they individually lack the neces-sary market power to influence oil prices.” If thischanges, this paper will have to be correspondinglyupdated.

Another caveat is that in this paper, we haveonly examined the historical relationship betweenthe price of WTI crude oil and EIA’s OPEC sparecapacity data, conditional on U.S. crude oil invento-ries. A future paper will examine this relationship,conditional on global inventory data.

Conclusion

At least from an examination of data over the past20 years, OPEC spare capacity has only matteredwhen (U.S.) crude oil inventories have been below

a threshold level. We would caveat our results bynoting this conclusion only has a practical use if thestates-of-the-world that occurred historically willcontinue to be the case going forward.

ABOUT THE AUTHOR

Hilary Till: Principal at Premia Research LLC, SolichScholar at the J.P. Morgan Center for Commodities at theUniversity of Colorado Denver Business and SchoolCo-Editor of Intelligent Commodity Investing. PremiaResearch LLC is a consulting firm that designs indices.Premia Research LLC starts with the premise that allmarkets can become fundamentally overstretched.Accordingly, an index should either include naturalhedges because of the potential of a market crash, or itshould dynamically allocate out of a market duringextremes in valuation. The design of the firm’s indicesreflects these beliefs.

Email address:till@premiarisk.com

ABOUT THE ARTICLE

Submitted: October 2015.Accepted: December 2015.

20energisk.org

References

[1] Büyüksahin B., M. Haigh, J. Har-ris, J. Overdahl and M. Robe. Funda-mentals, Trader Activity and Deriva-tive Pricing. EFA Bergen MeetingsPaper, December 4, 2009. Availableat http://ssrn.com/abstract=966692

[2] Büyüksahin B. The Price of Oil: Fun-damentals v Speculation and Data vPolitics IEA Oil Maret Report. SlidePresentation. 2011.

[3] Energy Information Administration(EIA). What Drives Crude OilPrices?. US Department of Energy,Presentation at Washington, D.C.January, 8 2014.

[4] Harrington, K. Crude Approxima-tions. Clarium Capital Managemnt,November. 2005.

[5] International Monetary Fund (IMF).Will the Oil Market Continue tobe Tight?.World Economic Outlook,Chapter IV, pp. 157-183. April, 2005.

[6] Kibsgaard, P. Scotia Howard Weil2015 Energy Conference. New Or-leans, Slide Presentation. March, 232015.

[7] Plante, M. and M. Yücel. Did Specu-lation Drive Oil Prices? Market Fun-damentals Suggest Otherwise. Fed-eral Reserve Bank of Dallas Eco-nomic Letter, Vol. 6, No. 11, October.2011.

[8] Till, H. Oil Futures Prices andOPEC Spare Capacity. Encana Dis-tinguished Lecture, J.P. MorganCenter for Commodities, Univer-sity of Colorado-Denver BusinessSchool, September 18th. 2014. Avail-able at: http://goo.gl/fRsuKS

[9] Till, H. Structural Positions in OilFutures Contracts: What are theUseful Indicators?. Argo Magazine,v. 6, Spring 2015. Available at:http://goo.gl/a21cvB

[10] Wright, B. The Economics of GrainPrice Volatility. Plenary Address at14th Annual Conference on GlobalEconomic Analysis. Slide Presenta-tion. Venice, June, 16. 2011.

Fall 201521

ADVERTISING FEATURE

http://ecomfin2016.essec.edu/

22

in the previous issue

Summer 2015 A

energy & commodity finance

Off-Shore Wind Investmentunder Uncertainty: Part II

OW Bunker: from IPO to Bankruptcy

banking & finance

Modelling of Libor-OIS basis

last issues are available at www.iasonltd.com and www.energisk.org